Loss of claudin-3 expression increases colitis risk by promoting Gut Dysbiosis

Figures & data

Figure 1. Robust decrease in CLDN3 expression characterize IBD patients and mice subjected to colitis.

(a) In silico analysis of the published IBD patient (n = 360) transcriptomics data demonstrating significantly downregulated CLDN3 expression versus normal (n = 153) individuals. ; (b) Immuno-histochemical analysis of CLDN3 expression in biopsy samples from IBD patients; (c) % weight change in mice weight when subjected to acute DSS-colitis (2.5% w/v); (d) Colon length; (e) Increased colon thickness (gm/cm) in DSS colitis; (f) Disease activity index; (g and h) Cumulative injury score and representative images from control and DSS colitis mice; (i) Immunoblot analysis of CLDN3 and ECAD in inflamed and control epithelial cells from colons. β-Actin was used as loading control; (j) Immunofluorescence analysis using anti-CLDN3 antibody in mice subjected to DSS colitis; (k) Change in % body weight in C57B/L6 mice subjected to C. rodentium infection; (l) Colon length; (m) Colon thickness in mice subjected to C. rodentium colitis; (n and o) Representative H&E images and injury score in mice subjected to C. rodentium infection; (p) Immunoblot analysis using anti-CLDN3, -ECAD and $\beta$-Actin antibodies infected with C. rodentium; and (q) Immunohistochemistry analysis using anti CLDN3 antibody in mice subjected to C. rodentium infection. *P < .05, **P < .01, ***P < .001 and ****P < .0001. Scale bar = 50 μm.

Figure 2. Cldn3KO mice are susceptible to colitis.

Cldn3KO mice and WT littermates were subjected to murine models of colitis to evaluate the role of colonic CLDN3 in IBD. (a) % change in body weight of Cldn3KO mice during acute DSS colitis; (b and c) Colon length and colon thickness in Cldn3KO mice versus WT littermates upon acute DSS treatment; (d) Disease activity index; (e and f) Representative H&E images and injury score in Cldn3KO mice subjected to acute DSS colitis; (g and h) Colon length and colon thickness in Cldn3KO mice subjected to chronic DSS colitis; (i and j) Representative H& E images and injury score in Cldn3KO (vs WT) mice subjected to chronic DSS colitis; (k) % change in body weight of Cldn3KO mice subjected to C. rodentium colitis; (l) Colon thickness in Cldn3KO mice versus WT littermates, and (m and n) Representative H&E images and collective injury score in Cldn3KO mice compared to WT littermates. *P < .05, **P < .01, and ***P < .001. Scale bar = 50 μm.

Figure 3. Deletion of Cldn3 expression exhibits deregulated inflammatory signaling.

Multiplex analysis was done using colon from naïve Cldn3KO and WT mice. Further, proinflammatory signaling was investigated in Cldn3KO subjected to murine models of colitis. (a) Cldn3KO mice exhibits dysregulation of cytokine level; (b and c) qRT-PCR analysis for IL-6 and TNF-α cytokines in the colon of Cldn3KO mice upon DSS and C. rodentium (vs WT); (d) Immunoblots analysis for STAT3 activation in Cldn3KO mice upon acute DSS and C. rodentium induced colitis; (e) Immunohistochemistry analysis of STAT3 activation in Cldn3KO mice compared to WT mice. *P < .05, **P < .01, ***P < .001 and ****P < .0001. Scale bar = 50 μm.

Figure 4. Pseudo-germ-free (microbial depletion) Cldn3KO mice show reduced severity of DSS-colitis.

Broad-spectrum antibiotic cocktail pretreatment was done to evaluate the role of gut microbiota in susceptibility of Cldn3KO mice to DSS colitis. (a) Experimental design for the administration of broad-spectrum antibiotics (Abx) and induction of DSS colitis; (b) Fecal microbial culture on LB-agar plate showing microbial depletion in Cldn3KO mice upon antibiotic treatment; (c) % body weight of antibiotics treated Cldn3KO mice upon DSS treatment compared to non-Abx-treated Cldn3KO and WT mice; (d and e) Colon length and thickness in Cldn3KO mice exhibits less effects of DSS treatment; (f) Disease activity index, and (g and h) Injury Score and representative H&E analysis reveal a reduction in disease severity in antibiotics treated Cldn3KO mice upon DSS colitis compared to non-Abx-treated Cldn3KO and WT mice. *P < .05, **P < .01, ***P < .001 and ****P < .0001. Scale bar = 50 μm.

Figure 5. Loss of endogenous Cldn3 expression alters gut microbial composition.

High throughput 16S DNA sequencing profiling analysis of fecal content was carried out using stool genomic DNA from naïve Cldn3KO and WT mice. (a and b) The distribution of frequency of features in each sample from Cldn3KO mice and WT controls; (c and d) Alpha rarefaction (observed and Shannon) curves show a lower diversity in WT littermates; (e) Principal coordinates analysis showing comparison of alpha diversity (Shannon) in Cldn3KO mice versus WT; (f) Principal coordinates analysis showing beta diversity (Jaccard) in Cldn3KO mice versus WT, and (g) Unweighted-Unifrac (PERMANOVA) analysis for beta diversity in Cldn3KO and WT mice.

Figure 6. Relative abundance of bacterial communities in Cldn3KO differ from WT.

Feces were collected for 16S rRNA gene sequencing and QIIME2 software was used to analyze the relative abundance of bacterial communities. Several bacterial taxons were altered with loss of Cldn3 compared to WT. Wilcoxon rank-sum test of bacterial core genera used in analysis between Cldn3KO mice compared to WT. (a) Relative abundance difference at phylum level as shown by the stacked bar; (b) Box plot showing lower abundance of Actinobacteriota in Cldn3KO mice (vs WT); (c) Increased Bacteroidota commensal in Cldn3KO mice; (d) Proteobacteria was found low in abundance in Cldn3KO compared to WT, and (e) Differential abundance of Firmicutes in Cldn3KO mice. *P < .05.

Figure 7. Gut dysbiosis in Cldn3KO mice leads to altered host metabolism.

Based on relative abundance of gut microbiota KEGG functional pathway analysis was done and significant metabolism pathways are presented. (a and b) Up and down regulated KEGG functional metabolic pathways.

Figure 8. Fecal microbiota transplantation (FMT) from Cldn3KO mice to germ-free C57BL/6 mice induces susceptibility to DSS colitis.

Fecal microbiota transplantation (FMT) was performed with germ-free mice to examine the effect of CLDN3 loss mediated dysbiosis on susceptibility to experimental colitis and/or severity. (a) Schematic representation of how FMT and DSS colitis was carried out; (b) % body weight of Cldn3KO FMT (vs WT-FMT) mice decreases upon DSS treatment; (c) Survival plot showing the death in Cldn3KO FMT mice when subjected to DSS colitis; (d) Cldn3KO FMT mice showed increased disease activity index compared to WT FMT; (e) Colon thickness increases in Cldn3KO FMT mice (vs WT FMT), and (f and g) Representative H&E images and injury score show the susceptibility in Cldn3KO FMT mice to DSS colitis *P < .05 and **P < .01. Scale bar = 50 μm.

Figure 9. Cldn3KO microbiota potentiates inflammatory signaling in intestinal epithelial cells.

Cldn3KO and WT fecal microbiota used to determine epithelial-microbial interactions dependent proinflammatory signaling and effect on mucosal barrier function. (a) Schematic representation how experiment was done using fecal microbial content; (b) Immunoblot analysis showing upregulated proinflammatory signaling in Cldn3KO fecal content treated cells than WT mice. (c and d) IL-6 and TNF-α production and secretion upon microbial content exposure to caco-2 cells; (e) Trans-epithelial resistance decreases in Cldn3KO treated microbial content versus WT; (f and g) Immunoblotting and densitometric analysis for CLDN3, and CLDN1, in Caco-2 cells exposed to fecal content from Cldn3KO and WT mice, and (h) Immunoblot analysis using anti-cyclin D1 and -Cleaved Caspase-3 in Caco-2 cells exposed to fecal content from Cldn3KO and WT mice. **P < .01, ***P < .001 and, ****, P < .0001.

Figure 10. Microbiome associated signaling is altered in IBD and corelates with CLDN3 expression.

Unbiased functional enrichment of DEGs obtained from in-silico meta-analysis of the previously published high throughput transcriptome from IBD patients was done using QIAGEN’s Ingenuity® Pathway Analysis (IPA) tool. Further interactions between DEGs and dysbiotic microbes were analyzed. (a and b) Canonical up and down regulated ingenuity pathway analysis using GSE database showing enrichment of pathways associated with bacterial signaling in IBD patients; (c) Disease and function analysis; (d) upstream regulator analysis in IBD patient cohorts; (e) Representative plots of Spearman correlation analysis of upstream regulator genes with CLDN3 expression.

Figure 11. Host-microbiome interactions in IBD patient cohort show overlap with Cldn3KO mice.

Comparative analysis was performed to determine overlaps in microbial communities between IBD patients and Cldn3KO mice, the vital gene interactions with bacteria, and the correlation with CLDN3 Expression. (a) Representative image showing the overall overlap between host genes: microbiota interaction between human IBD (vs normal) and Cldn3KO (vs. WT) mice; (b) Interaction among gene expression in IBD and dysbiotic OTUs from IBD patient and Cldn3KO mice; (c) Correlation analysis of genes with CLDN3 Expression, and (d) Schematic representation of overall concept.

Data availability statement

The data are available upon request. The data supporting the findings of this study are available from the corresponding author upon request.